The present invention relates to a binary-coding system for the conversion of an analog picture signal read out by an image sensor into a digital signal.
In a reader which reads out two-dimensional picture information with a one-dimensional image sensor, an analog picture sigal read out by the image sensor is subjected to binary-coding processing to be in correspondence to the lightness (black-and-white) of an original at each main scanning line. FIG. 1 shows a conventional binary-coding system, wherein an analog picture signal 1 is compared with a reference voltage 3 by a comparator 2, thus obtaining a binary-coded signal 4.
However, when the analog picture signal 1 having variations in levels according to a difference in thickness and/or density of characters on the original is supplied as shown in FIG. 2A, a binary-coded signal 4 (FIG. 2B) missing considerable amount of picture information will be produced from such a simple binary-coding system no matter how the reference voltage 3 is set. Therefore, an improved binary-coding system has been recently employed in which an analog picture sigal is differentiated to emphasize a change point prior to implementing the binary-coding processing. According to this improved binary-coding system, the reproducibility of the picture signal read out by the image sensor is improved in the main scanning direction. However, this system is not capable of improving reproducibility in the subscanning direction, and thus the aforementioned problem in picture quality (namely, the deterioration produced by the binary-coding of an analog picture signal) still remains.